Principle and design of FRS for the treatment of OMFig. 1

Illustration and property of FRS hydrogels in the treatment of OM disease. (a) Schematic illustration of FRS in the ME. Traditional strategy is usually accompanied by a fluid drainage of antibiotic aqueous solution through TM or ET. CS hydrogel cannot stay in adequate contact with bacterial colonies and biofilms. FRS hydrogel in this work eliminates multiple bacterial, especially antibiotic-resistant bacteria, in the enclosed TC and restores its hearing. (b) Representative photographic of normal CS hydrogel and FRS hydrogel on the curved ME mucosa, showing the existence of air gaps and conformal topology, respectively (scale bar: 5 mm, 500 µm). (c) Schematic illustration of the phase transition mechanism of the FRS hydrogels. (d) Plot of the reported antimicrobial biological additives based on their price of raw materials and reported antibacterial rate. FRS, inorganic copper sulfide (CuS), silver nanoparticles (Ag NPs), molybdenum disulfide nanosheets (MoS2), MXene, polypeptide ε-polylysine, antibiotics OFL, ionic liquid. (e) Photographs and infrared camera images of FRS hydrogels at body temperature (37℃) and low temperature (4℃), demonstrating the transition between fluidic state at low temperature and solid-like gel state at body temperature (scale bar: 2 cm). (f) The viscosity and (g) the calculated tanδ (G’’/G’) of FRS hydrogels with different recipes at a temperature sweep from 0℃ to 55℃, showing the sol-gel transition

Ordinarily, the use of OFL ear drops prevails in clinical practice due to their potent antimicrobial efficacy and cost-effectiveness. However, as depicted in Fig. 1a, the high fluidity of OFL leads to the leakage through the tympanic membrane (TM) or eustachian tube (ET), resulting in suboptimal bactericidal efficacy and a heightened recurrence rate. Moreover, conventional injectable non-FRS hydrogels often exhibit inadequate conformal ability, forming numerous air gaps between the gel and TC mucosa. Hence, we opted for stimuli-responsive FRS hydrogels comprising FDA-approved CS and Pluronic F127 as injectable antibacterial agents for OM treatment. These innovative hydrogels facilitate the effective delivery of antibacterial agents into the TC, transitioning between a flowing injectable state in vitro and a non-flowing state in vivo, thereby ensuring prolonged and conformal contact with the bacterial colony within the TC mucosa, especially in intricate areas, and complete biodegradation and assimilation in vivo.

The FRS hydrogel showcases exceptional topological conformal properties attributable to its in vivo phase transition, enabling it to flow into and adhere to bumpy corners and TC mucosa. To validate this feature, FRS hydrogel was applied to curved pig epidermis at a low temperature to mimic ear conditions. As illustrated in Fig. 1b, numerous air gaps were observed between the CS hydrogel and the curved epidermis, whereas FRS tightly adhered to the surface due to its phase transition characteristics. Similarly, FRS hydrogel exhibited superior fluidity, enabling it to penetrate tiny corners of a customized mold, unlike non-FRS CS hydrogel, which faced limitations due to high viscosity (Fig. S1a).

The FRS hydrogel formulation primarily consists of thermosensitive Pluronic F127 and biocompatible antibacterial components, which undergo gelation in vivo due to the near-body temperature lower critical solution temperature (LCST) of the Pluronic F127 triblock copolymers (Fig. 1c). While various antibacterial additives have been explored recently [24,25,26,27,28], including inorganic nanoparticles and organic polypeptides, we advocate for CS due to its high antibacterial efficacy (approximately 90%), widespread availability, and cost-effectiveness (0.016 $/mL), along with its broad-spectrum antibacterial properties against Gram-positive, Gram-negative, and antibiotic-resistant strains. In contrast, emerging alternatives like MXene, polypeptides, and Ag NPs, which have similar antibacterial efficiencies, involve significantly higher costs [14, 29,30,31,32,33,34,35,36,37,38,39,40,41,42,43]. Additionally, commonly used clinical ofloxacin ear drops (purchased from a pharmacy in Shanghai, China) are priced at 0.63 $/mL, which is eight times higher than the cost of FRS (0.07 $/mL), as shown in Fig. 1d and Fig. S3.

The tunable fluidity was assessed through rheological analysis. FRS hydrogel transitions from a fluidic state at low temperatures to a gel-like consistency at body temperature, facilitated by reversible alterations in hydrophobicity and intermolecular hydrogen bonding interactions [29], showcasing thermally responsive and reversible sol-gel transition properties (Fig. 1e). Different concentrations of CS in FRS hydrogels modulate their fluidity, impacting topological conformal properties (Fig. S1b). The gelation time of FRS hydrogels in the TC was within minutes and varied based on CS content, with higher concentrations impeding fluidity and antibacterial efficacy within the enclosed TC (Fig. S1c, S1d), as evidenced in Fig. S2. The rheological properties of various FRS hydrogels were thoroughly analyzed in response to temperature fluctuations to confirm the temperature-regulated reversible sol-gel transition (Fig. 1f and g, S1e). Except for the neat CS hydrogel, all mixed hydrogels exhibited temperature-sensitive changes in storage modulus (G’) and loss modulus (G’’), as evidenced by a decrease in tanδ (G’’/G’) to below 1 and an increase in loss modulus with rising temperatures. These rheological characteristics enable the FRS to be administered at room temperature by dripping the liquid formulation through the external auditory canal, where it naturally flows into the middle ear cavity under the influence of gravity, achieving uniform distribution due to its excellent fluidity. Upon reaching body temperature, the hydrogel undergoes a sol-to-gel phase transition, ensuring positional stability and preventing leakage from body movements, thereby providing prolonged therapeutic effects.

Therapeutic efficacies of FRS hydrogels on antibiotic-resistant OM treatmentFig. 2

The therapeutic effect of FRS hydrogels on antibiotic-resistant OM. (a) Schematic representation showing the procedure of the establishment and treatment of antibiotic-resistant OM rat. (b) Schematic diagram and representative photographic images of the recovered TMs at different degrees of closure and transparency at day 7 (scale bar: 500 μm), cultured bacteria at day 7 by washing the surgical removed ear bullae with saline (scale bar: 1 cm), and calculated thickness of TMs (scale bar: 200 μm) and ME mucosa (scale bar: 200 μm) at day 7 of different groups, including CON, F127, OFL, and two FRS groups. (c) The corresponding statistical data according to the above photos, including the observed healing days required for the closure of perforated TMs; the quantitative score of recovered TMs at day 7 in terms of the observed TMs closure, vascular proliferation, secretion, TMs thickening and fibrosis; the statistic bacteria viability in TC at day 7; the counted TM and mucosal thickness of different groups at day 7. (d) Immunofluorescence staining of CD206 (green), CD86 (red), and nuclei (blue) of untreated OM rats (CON) and 1%FRS-treated rats (scale bar: 200 μm), and their corresponding quantification of the ratio of CD206 to CD86. (e) The anti-bacterial and (f) anti-biofilm properties of different groups against OFL-resistant bacteria in vitro, including CON, F127, OFL, FRS. Fluorescence staining (scale bar: 100 μm) and spread-plate culture photos after 1 day (scale bar: 1 cm) of residual OFL-resistant S. aureus treated by different groups. The variation of OFL-resistant S. aureus biofilm by stained with crystal violet (scale bar: 2 mm), or propidium iodide (PI) and Syto-9 (scale bar: 100 μm). (g) The statistical chart and image of inhibition zone of OFL and 1%FRS groups (scale bar: 1 cm). (h) Representative SEM images of OFL-resistant S. aureus morphologies before and after FRS treatment (scale bar: 1 μm), where shrunk membrane can be observed. (i) The schematic diagram of bacteria aggregation experiment and its calculated detached bacterial of different groups. All statistical analyses were performed by one-way ANOVA, data represent mean ± standard deviation, *p < 0.05, **p < 0.01, ***p < 0.001

Antibiotic resistance poses a significant challenge to effective antibacterial therapy in the present era, diminishing bacteriostatic efficacy. This issue is particularly pronounced in the treatment of OM, where the ototoxicity of aminoglycosides, tetracyclines, and other antibiotics has limited the available treatment options. To address this, we conducted a comprehensive evaluation of FRS’s antibacterial efficacy against antibiotic-resistant bacteria in vivo (Fig. 2). OFL-resistant S. aureus was selected as the pathogen due to the widespread use of OFL in clinical OM management. Control groups included a blank control (CON), 1% w/v CS, and 0.3% w/v OFL, while FRS hydrogels with varying CS concentrations (0.1%, 1%, and 2% w/v) and FRS-OFL with 0.3% w/v OFL were also examined. Notably, the impact of conformal contact was assessed by comparing the healing outcomes of the OFL and FRS-OFL groups, as well as the CS and 1%FRS groups.

An OFL-resistant OM model was established by injecting OFL-resistant S. aureus into the ME (Fig. 2a), with therapeutic efficacy evaluated through quantitative assessments of TM healing, residual bacteria in the ME, and TM and mucosal thickness across different groups (Fig. 2b and c). Otoscopy was employed to visually assess TM recovery, revealing that FRS significantly expedited TM healing post OFL-resistant S. aureus infection. Notably, the perforated TM treated with 1%FRS displayed transparent closure by day 7, devoid of significant tissue proliferation. In contrast, TMs treated with CS or OFL exhibited persistent perforation or closure with evident inflammation, edema, and suppuration, starkly differing from the normal lamina propria structure. The FRS-OFL treated TM showcased transparent closure with radial vascularity and congestion. Hematoxylin and eosin (H&E) staining further confirmed the superior healing outcomes of 1%FRS, as evidenced by a significantly reduced TM thickness (17.63 ± 7.44 μm) compared to CS (67.65 ± 18.32 μm), OFL (48.70 ± 9.12 μm), and FRS-OFL (34.36 ± 4.77 μm) groups, aligning more closely with normal rat TM thickness (10.08 ± 6.39 μm). Additionally, we assessed the condition of the observed TM based on four parameters: TM closure, vascular proliferation, secretion, and TM thickening and fibrosis, with a scoring system where 0 denotes a return to normal, 1 signifies incomplete recovery, and 2 indicates no significant improvement. Notably, the healing outcomes for the 1%FRS group (8.6 ± 1.7 healing days, severity score of 1.0 ± 0.71) were markedly superior to those of the CS (13.6 ± 1.8 healing days, severity score of 6.2 ± 1.10), OFL (12.8 ± 2.0 healing days, severity score of 3.6 ± 1.14), and FRS-OFL (12 ± 2.2 healing days, severity score of 2.6 ± 1.14) groups. These findings underscore the optimal healing condition achieved with 1%FRS treatment of the TM.

Subsequently, bacteria were extracted from the ME under sterile conditions to quantify colony-forming units (CFU). Remarkably, 1%FRS exhibited the lowest colony count (0.12 ± 0.05 *105 CFU), while CS (1.57 ± 0.19 *105 CFU), OFL (0.76 ± 0.16 *105 CFU), and FRS-OFL (0.60 ± 0.15 *105 CFU) groups displayed approximately 13, 6, and 5 times higher CFU counts, respectively. The antibacterial efficiency of 1%FRS and OFL against OFL-resistant S. aureus, calculated based on CFU differences relative to the CON group, stood at 93.8% and 60.8%, respectively, underscoring the superior bacteriostatic properties of 1%FRS against antibiotic-resistant OM.

Furthermore, surgical removal of rat bullae enabled the observation of inflammatory responses. Notably, the ME mucosa in untreated, OFL, and CS groups exhibited marked edema and thickening (Fig. 2b), accompanied by increased inflammatory cell infiltration. In contrast, FRS-OFL and 1%FRS groups displayed mucosal characteristics closer to normal, with intact epithelium and reduced edema. Quantitative analysis further revealed that the mucosa in the 1%FRS group (32.43 ± 7.44 μm) was close to the normal rat mucosal thickness (21.5 ± 3.53 μm) [30], when those of CS (319.05 ± 30.72 μm), OFL (151.26 ± 32.19 μm) and FRS-OFL (80.09 ± 25.35 μm) were highly edematous, indicating superior therapeutic effects in OFL-resistant rats.

It is noteworthy that the bacteriostatic activity of the 2%FRS group was less effective than the 1%FRS groups in vivo (Fig. S4a and S4b), despite CS being the primary bactericidal agent in the mixed gel. The healing days (13.6 ± 1.14 days), TM thickness (57.81 ± 8.73 μm), the severity score of recovered TMs (5.4 ± 1.1), residual bacteria in ear (0.99 ± 0.20 *105 CFU) and mucosa thickness (109.44 ± 19.79 μm) of 2%FRS, were generally worse than those of 1%FRS. The inferior outcomes of the 2%FRS group were attributed to its increased viscosity (3324 mPa·s, almost 4 times higher than that of 1%FRS), which hindered topological conformal properties in the TC, emphasizing the critical role of hydrogel fluidity in antibacterial performance in vivo.

The inflammatory response of ME mucosa at day 7 was characterized by quantifying the M2/M1 macrophage ratio through immunofluorescent staining. CD206, a M2 macrophage marker shown in green, of 1%FRS group was clearly more than that of CON group (Fig. 2d), indicating the presence of more reparative M2 phenotypes. Notably, the 1%FRS group (2.54 ± 0.64) exhibited a significantly higher M2/M1 ratio than the CON group (0.63 ± 0.10), indicative of a more reparative inflammatory response.

To further explore the inhibitory effects of 1%FRS on the proliferation of OFL-resistant S. aureus, antibacterial properties and biofilm integrity were assessed in vitro by 6 h incubation and subsequently direct live/dead staining (direct fluorescence method) or diluted and incubated for an additional 24 h (spread-plate method). As shown in Fig. 2e, 1%FRS demonstrated superior bacteriostatic behavior against antibiotic-resistant S. aureus compared to OFL and F127. As calculated in Fig. S4c and S4d, the antibacterial efficiencies of 1%FRS were notably higher (73.48 ± 3.30% of direct fluorescence method and 69.38 ± 7.19% of spread-plate method) than those of OFL (59.76 ± 15.11%, 37.01 ± 11.00%) and F127 group (53.96 ± 9.52%, 12.32 ± 0.43%), underscoring its excellent bacteriostatic properties.

Eradication of bacterial biofilms is critical to prevent recurrence of OM. The antibiofilm properties of FRS against OFL-resistant S. aureus were confirmed through crystal violet staining (Fig. 2f), demonstrating the disruption of biofilms. The 3D reconstructions of live/dead staining of the bacterial biofilms were stained by Syto9/PI and observed by confocal laser scanning microscopy (CLSM), where the red fluorescence suggested an increase in the dead bacteria of different groups. Through statistics (Fig. S4e and S4f), crystal violet stained biofilm masses in 1%FRS (0.38 ± 0.09) were much less than that of OFL (0.86 ± 0.04), and fluorescence intensity of Syto9/PI stained dead bacteria in 1%FRS (73.83 ± 7.21) was approximately 6 times higher than that of OFL (11.66 ± 3.58), highlighting the efficacy of 1%FRS in disrupting OFL-resistant S. aureus biofilms.

Additionally, the inhibition zone of 1%FRS and OFL against OFL-resistant S. aureus was assessed (Fig. 2g), with 1%FRS (2.56 ± 0.53 cm2) exhibiting a significantly larger inhibition zone area than that of OFL (0.84 ± 0.06 cm2). SEM imaging in Fig. 2h further revealed the altered morphology of OFL-resistant S. aureus post-treatment with 1%FRS. Moreover, bacterial aggregation assays highlighted the enhanced capacity of different hydrogels to absorb OFL-resistant S. aureus, with 1%FRS displaying superior adsorption capabilities. After conversion (Fig. 2i), the ratio of residual OFL-resistant S. aureus in 1%FRS solution at 24 h (11.17%), significantly lower than those of F127 (55.82%) and CON groups (61.64%), demonstrating the excellent adsorption capacity of 1%FRS.

Accelerated therapy of OM by conformal contactFig. 3

Accelerated therapy of OM by FRS in vivo. (a) Representative images of recovered TMs, cultured residual bacteria, ME mucosa of different treatment groups, including CON, CS, OFL and two FRS groups. Photos of recovered TMs in different degrees of closure and transparency (scale bar: 500 μm), as well as different thickness (scale bar: 200 μm), cultured residual bacteria in ear by washing the surgical removed ear bullae with saline (scale bar: 1 cm), and H&E staining of the mucous membrane lining the ear bulla (scale bar: 200 μm) were shown. (b) Schematic representation showing the procedure of the establishment and treatment of OM rat. (c) The observed number of days required for the closure of perforated TMs, and the quantitative score of recovered TMs at day 7 in terms of the observed TMs closure, vascular proliferation, secretion, TMs thickening and fibrosis. (d) The corresponding statistic residual bacteria viability at day 7. (e) The corresponding statistic mucosal and TM thickness of different groups at day 7. (f) Immunofluorescence staining of CD206 (green), CD86 (red), and nuclei (blue) of untreated and 1%FRS treated rats, and (g) their corresponding quantification of the ratio of CD206 to CD86. (h) Immunohistochemistry staining of IL-1β, IL-6 in the ME mucosa of different treatment groups and (i) their corresponding statistic expression. All statistical analyses were performed by one-way ANOVA, data represent mean ± standard deviation, ns > 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

To further illustrate the enhanced effectiveness of conformal contact in the treatment of OM, we established an acute OM model using S. aureus (Fig. 3b). Images of TMs in Fig. 3a and S5a reveal that TMs treated with CS still exhibited significant pathological changes, while those treated with OFL were mostly healed but with accompanying fibrosis, thickening, and vascular proliferation. In contrast, TMs treated with FRS-OFL and 1%FRS showed almost complete recovery, with only slight residual fibrosis at the perforation site. Statistically, as show in Fig. 3c and S5b, the average healing time of the FRS groups (7.6 ± 1.5 days for FRS-OFL, 8.0 ± 2.0 days for 1%FRS) was significantly shorter than that of the non-FRS groups (11.6 ± 2.0 days for OFL, 14.6 ± 2.7 days for CS), as well as the F127 group (14.8 ± 2.5 days). The TM thickness in the FRS groups (14.07 ± 4.46 μm for FRS-OFL, 21.56 ± 13.02 μm for 1%FRS) was also lower than that in the control groups (36.04 ± 8.55 μm for OFL, 102.46 ± 17.45 μm for CS), shown in Fig. 3e. The quantitative severity score of TMs treated with FRS-OFL (1.2 ± 0.4) and 1%FRS (1.2 ± 0.8) was notably lower than that of OFL (3.2 ± 0.8) and CS (5.2 ± 0.8), indicating the role of FRS in accelerating TM healing.

Following that, we isolated bacteria from the treated ME. On blood agar plates, the FRS groups displayed minimal colonies on day 7, whereas the CON, CS, and OFL groups still had a substantial number of colonies visible. As illustrated in Fig. 3d and S5c, 1%FRS (0.27 ± 0.15 *105 CFU on day 5, 0.05 ± 0.07 *105 CFU on day 7) and FRS-OFL (0.51 ± 0.26 *105 CFU on day 5, 0.14 ± 0.12 *105 CFU on day 7) exhibited the fewest colonies, while OFL (2.11 ± 1.11 *105 CFU on day 5, 0.24 ± 0.25 *105 CFU on day 7) and CS (5.86 ± 2.31 *105 CFU on day 5, 1.02 ± 0.59 *105 CFU on day 7) were nearly 5 and 10 times higher, respectively, underscoring the superior antibacterial efficacy of FRS in vivo.

H&E staining of the ME mucosa revealed that the ME mucosa affected by OM was significantly edematous (266.81 ± 48.60 μm). Similarly, the ME mucosa thickness in the 1%FRS group (36.56 ± 13.42 μm) was notably thinner than that in the CS group (261.86 ± 56.12 μm), as well as in the OFL group (116.54 ± 13.51 μm) and FRS-OFL group (62.19 ± 19.98 μm). Furthermore, the analysis of the cell per unit area of the ME mucosa, as depicted in Fig. S5d, indicated a significantly lower cell count in the 1%FRS group compared to the control group, suggesting reduced inflammatory proliferation and exudation. As shown in Fig. S6, 2%FRS was also found to be less effective than 1%FRS in vivo. Parameters such as TM healing days (11.4 ± 0.4 days), quantitative score of recovered TMs (4.2 ± 1.1), residual bacteria in the ear (0.23 ± 0.14 *105 CFU on day 7), TM thickness (79.88 ± 9.37 μm), and mucosa thickness (101.34 ± 15.28 μm) were generally poorer than those of 1%FRS.

The inflammatory response of the ME mucosa was characterized by quantitatively analyzing macrophages and inflammatory factors. Macrophages, crucial immune cells influencing wound healing, can transition into proinflammatory M1 or reparative M2 phenotypes to regulate inflammation or promote tissue repair, respectively. Immunofluorescence analysis and quantitative results in Fig. 3f and g, and S7 indicated that the 1%FRS group had a higher M2/M1 ratio (2.15 ± 0.76) compared to the control group (0.30 ± 0.07), suggesting a higher proportion of M2 macrophages that can prevent the progression from acute OM to chronic OM, signifying recovery in the FRS group while inflammation persisted in the CON group. Furthermore, the expression of inflammatory factors analyzed through immunohistochemistry (Fig. 3h and i) demonstrated that the levels of interleukin-1β (IL-1β) and interleukin-6 (IL-6) in the 1%FRS group (0.14 ± 0.01, 0.16 ± 0.02) and FRS-OFL group (0.10 ± 0.03, 0.12 ± 0.02) were significantly lower than those in the OFL group (0.20 ± 0.05, 0.21 ± 0.03) and CON group (0.48 ± 0.12, 0.43 ± 0.13), indicating that the rats treated with 1%FRS had markedly lower inflammation levels and comparable recovery to conventional OFL ear drops treatment.

FRS could be metabolized in vivo, which was demonstrated through observation of TC after FRS injection. As shown in Fig. S8a and b, FRS hydrogel was completed dissolved in PBS at 37℃, and no FRS was observed in the surgically removed TC. Intraepidermal injection was used to watch the metabolic processes of FRS (Fig. S8c). Over time, the bulge was gradually diminished, suggesting the absorption of FRS in vivo.

Therapeutic efficacies of FRS on chronic OM ratsFig. 4

Optimum therapeutic efficacy of 1%FRS on chronic OM rat. (a) Schematic representation showing the procedure of the establishment and treatment of chronic OM rat. (b) Representative images of the infected TMs and residual bacteria in ME on day 42 and 56 after treated by CON, OFL, FRS-OFL and 1%FRS. (c) The estimated disease severity score of healed TMs and (d) calculated residual bacteria viability in ear of different groups. (e) Schematic diagram of the chronic OM rat with left ear untreated and right ear treated with 1%FRS, resulting in asymmetric vestibular dysfunction. (f) A set of photos showing the anti-clockwise spin of above chronic OM rat in seconds. (g) Micro-CT images showing the 3D reconstructions of the chronic OM rat brain with recovered right ear and damaged left ear at day 56. All statistical analyses were performed by one-way ANOVA, data represent mean ± standard deviation, ***p < 0.001

The therapeutic efficacy of antibacterial drugs on a chronic OM model holds significant clinical relevance due to the challenge of eradicating persistent bacterial colonies. In this study, we established a chronic OM model following a previously reported protocol (Fig. 4a) [31]. As anticipated, similar treatment trends to those observed in acute OM were noted (Fig. 4b). Rats treated with 1%FRS and FRS-OFL exhibited no evident suppuration in the TM on days 42 and 56, indicating lower infection rates in the FRS groups. In contrast, rats in the non-FRS CON and OFL-treated groups displayed yellowing of the TM, along with severe hypertrophic scarring and purulent inflammation. Notably, in the CSOM model, we observed spontaneous polymicrobial infections (Fig. 4b). On day 14 of FRS treatment, the bacterial culture results from the middle ear revealed the presence of various colonies in all treatment groups. In the untreated control group, the predominant colonies were those of P. aeruginosa (used in the model, appearing as flat green colonies) with a few S. aureus colonies (appearing as small white dots). In the FRS-OFL treatment group, S. aureus was predominant, with a few P. aeruginosa colonies present. However, in the OFL and 1% FRS groups, no P. aeruginosa colonies were observed, and the number of S. aureus colonies was significantly reduced. This indicates that during the course of chronic otitis media, the rats experienced secondary bacterial infections. In untreated rats, both the primary infection was severe and persistent, and the secondary infections were relatively serious. In contrast, 1% FRS effectively limited the primary bacterial infection, protecting the middle ear tissue and significantly reducing the severity of secondary bacterial infections. Consequently, the disease severity scores of 1%FRS (2.8 ± 0.8) and FRS-OFL (3.4 ± 0.5) were notably lower than those of CON (7.2 ± 0.8) and OFL (4.8 ± 0.8), as were the residual bacterial viabilities (0.56 ± 0.15 *104 CFU for 1%FRS, 0.67 ± 0.25 *104 CFU for FRS-OFL, 36.19 ± 9.98 *104 CFU for CON, 3.99 ± 1.04 *104 CFU for OFL), shown in Fig. 4c and d. It is important to highlight that the therapeutic efficacy of 1%FRS surpassed that of the clinically used OFL, underscoring the significant potential of FRS hydrogels in the clinical management of OM.

The vestibular function of the treated rats was evaluated, as prolonged and recurrent infections in chronic OM rats not only led to ME inflammation but also resulted in vestibular function impairment, causing balance issues in the rats. As depicted in Fig. 4e, to visually demonstrate the therapeutic benefits of FRS hydrogels in chronic OM, we treated the right ear of a chronic OM rat with 1%FRS, leaving the left ear untreated as a control. Intriguingly, behavioral analysis revealed that this chronic OM rat exhibited anti-clockwise spinning in the absence of external stimuli (Fig. 4f), indicating severe unilateral vestibular organ damage in the untreated left ear and restored vestibular function in the right ear (see Movie S1). Additionally, to assess the severity of lesions in the ME, micro-computed tomography (micro-CT) scans were performed on normal rats and chronic OM rats. The results in Fig. S9 demonstrated that the temporal cavity of normal rats appeared hollow, whereas hypodense shadows (indicated by white arrows) were observed in the temporal cavity of chronic OM rats, suggesting the presence of residual fluid accumulation. Furthermore, as shown in Fig. 4g, the affected left ear of the chronic OM rat with vestibular dysfunction exhibited substantial fluid accumulation (white arrow) and noticeable structural damage to the inner ear (red arrow), while the contralateral right ears treated with 1%FRS remained structurally intact. The micro-CT and behavioral findings corroborated that 1%FRS effectively impeded the progression and deterioration of chronic OM to a certain extent, preventing further damage.

ABR test of FRS-treated OM ratsFig. 5

Auditory brainstem response (ABR) test of the FRS treated rats. (a) Schematic representation of neurons in the auditory nerve and brainstem firing in response to evoked acoustic stimulus activity, and photos of the ABR experimental set-up. (b) Representative ABR waveforms evoked by click of different rats under acoustic stimulation. (c) The calculated ABR threshold of untreated and 1%FRS treated rats at day 14 under different frequencies stimulation. (d) The measured ABR wave I amplitude of untreated and 1%FRS treated rats. (e) ABR waveform evoked by click of untreated and 1%FRS treated rats at day 7, and (f) their measured latency. (g, h) Comparison of ABR waveforms evoked by click and their threshold of healthy and 1%FRS treated rats at day 14 and day 28, showing the hearing recovery of FRS treated rats

To assess the hearing recovery and biocompatibility of FRS in the ear, we conducted auditory brainstem response (ABR) tests to evaluate the hearing of FRS-treated rats. ABR is a type of electroencephalogram signal that responds to auditory stimuli and is commonly used to assess potential auditory function disorders within the brain. The experimental setup is illustrated in Fig. 5a, where rats were positioned in an anechoic chamber to eliminate electromagnetic and acoustic interference during the experiments. Further details on the experimental procedure and setup can be found in the Supporting Information. Representative waveforms of an untreated OM rat, an OM rat treated with FRS, and a healthy rat in response to clicks (100 µs duration) are depicted in Fig. 5b, with the black lines denoting the specific sound pressure level (SPL) of the acoustic stimulation, revealing discernible peaks characteristic of regular III and V responses. The corresponding calculated ABR thresholds are presented in Fig. 5c and S10, where the threshold of FRS-treated rats, particularly the 1%FRS group, was significantly lower than that of untreated OM rats for both click and all pure tone stimuli (1, 2, 4, 8, 16, and 32 kHz) on day 14. Notably, no statistical difference in ABR thresholds was observed between normal rats and those treated with FRS, indicating successful hearing recovery in FRS-treated rats.

To further elucidate the improvement in hearing, we analyzed ABR wave I amplitudes, reflecting the number of firing neurons, and ABR wave I latencies, indicating the speed of transmission. At day 7, FRS-treated rats exhibited significantly higher ABR wave I amplitudes at frequencies of 4, 8, and 16 kHz compared to untreated OM rats (Fig. 5d). Moreover, the amplitude of the representative waveform of FRS-treated OM rat evoked by click was notably higher than that of untreated OM rat (Fig. 5e). Similarly, shorter ABR wave I latencies in FRS-treated rats compared to untreated OM rats suggested a reduced delay in cochlear nerve transmission (Fig. 5f) [32]. In Fig. S10b, the waveforms of untreated and FRS-treated OM rats displayed higher amplitudes and lower latencies in the FRS group, indicating improved hearing. Collectively, these findings demonstrate an enhancement in hearing in FRS-treated OM rats.

To assess whether the hearing has returned to normal, a comparison of ABR results between healthy rats and FRS-treated rats is presented in Fig. 5g and h. The ABR threshold of 1%FRS-treated rats was nearly equivalent to that of healthy rats after 14 days, although differences in wave amplitude and latency persisted. However, by day 28, the hearing data were nearly identical, indicating complete restoration of hearing in 1%FRS-treated rats. These results suggest that FRS is non-ototoxic, providing an additional advantage over antibiotics.

The cytocompatibility of FRS hydrogels was evaluated in Fig. S11 through live/dead assays of mouse fibroblast L929 cells, primarily utilizing 2%FRS due to its higher content. As depicted in Fig. S11a-c, no significant differences in viable cells were observed between the CON and FRS groups after a 24-hour incubation period, indicating that FRS hydrogels did not impact cell growth and proliferation. Regarding the potential long-term risks associated with FRS, previous clinical studies have demonstrated the safety of CS as an oral medication in humans [44]. Therefore, we further investigated whether FRS has any adverse effects on the structure of middle ear tissues, we performed micro-CT scans on rats two months after FRS treatment (Fig. 4g, S9). The results demonstrated that the middle ear structure of the treated ears remained intact and healthy. To assess the potential allergic risk associated with FRS, we conducted RNA-seq analysis on middle ear tissues from rats treated with FRS (Fig S11d). The results showed no statistically significant differences in the expression of genes related to allergic reactions compared to the control group, indicating a low risk of allergic responses induced by FRS. Furthermore, to assess biocompatibility in vivo, as shown in Fig. S11e, rats injected with 2%FRS for 7 days were euthanized to obtain tissue samples from major organs (heart, liver, spleen, lung, and kidney). Histopathological examinations from H&E staining sections revealed no significant pathological damage induced by FRS injection in the major organs, indicating the absence of organ toxicity associated with FRS treatment.

Bacteriostatic activity of FRS hydrogels against multiple pathogenic bacterial of OMFig. 6

Antibacterial and antibiofilm properties of FRS hydrogels in vitro. (a) Fluorescence staining (scale bar: 100 μm) and spread-plate culture photos (scale bar: 1 cm) of common clinical OM pathogenic bacterial treated by different groups, including Gram-positive S. aureus, S. pneumoniae, and Gram-negative P. aeruginosa, E. coli. The color of the S. pneumoni